Magnetic shielding barrel noise floor evaluation method based on cross-spectrum analysis

Abstract

A magnetic shielding barrel noise floor evaluation method based on cross-spectrum analysis comprises the steps of data acquisition, simultaneous acquisition of noise data at the central position of a magnetic shielding barrel by using two independent atom magnetometers, windowing processing of discretized noise data of two channels by using a Hanning window, and reduction of near side lobe leakage. The method comprises the following steps: windowing two-channel data, performing Fourier transform on the windowed two-channel data to obtain frequency spectrum data, and finally performing frequency domain cross-correlation conjugate operation to calculate frequency domain characteristics of common-mode components of the two atom magnetometers and obtain a noise spectrum of the common-mode components of the two atom magnetometers, namely the noise floor of the magnetic shielding barrel. According to the method, common-mode components between the channels are extracted through cross-spectrum analysis operation, so that the noise floor of the magnetic shielding barrel or the noise floor of the magnetic shielding environment is evaluated, a metering type magnetometer with higher precision does not need to be researched, developed and used, the noise floor of the magnetic shielding barrel can be evaluated only through two magnetometers with lower sensitivity, the cost is lower, and the method is suitable for popularization and application. The equipment complexity of the extremely weak magnetic measurement device does not need to be increased, and the credibility is high.

Description

Technical Field

[0001] The present invention relates to the field of magnetic field measurement technology and weak signal detection, and in particular to a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis. Background Art

[0002] The measurement of extremely weak biomagnetic fields is of great significance in medical surgery and disease diagnosis. In order to achieve the measurement of brain magnetism at the order of hundreds of fT and heart magnetism at the order of hundreds of pT, the extremely weak magnetic field measurement device needs to rely on a magnetic shielding device to shield the strong geomagnetic field and various electromagnetic equipment. Therefore, the performance of the magnetic shielding device is directly related to the sensitivity of the core sensor in the biomagnetic measurement device, the atomic magnetometer. As the main shielding device currently used in heart and brain magnetism measurement, the shielding barrel is usually composed of four layers of permalloy shielding layers to shield the low-frequency magnetic field, and then constructed with pure aluminum materials to shield the high-frequency magnetic field. The key indicators of the magnetic shielding barrel include the shielding factor and the noise floor inside the shielding barrel. Due to differences in shielding materials, structures, and processes, the shielding barrel indicators are directly affected by them, and the height of the shielding environment noise floor determines whether it is the main noise interference source of the biomagnetic measurement device. Therefore, it is very important to measure the noise of the shielding barrel.

[0003] At present, the central noise floor of the fully enclosed magnetic shielding barrel designed and manufactured in China can reach 10fT / Hz -1 / 2 Below, the internal combination of Permalloy and ferrite barrel can even reach as low as 1fT / Hz -1 / 2 The sensitivity of the most sensitive commercial magnetic measurement sensor at home and abroad is 30fT / Hz. -1 / 2 , it is difficult to directly evaluate the shielded barrel noise, so it is necessary to develop a more accurate metrological magnetometer. Summary of the invention

[0004] In view of the above technical problems, the purpose of the present invention is to provide a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis, which can be achieved by using 30fT / Hz -1 / 2 Sensitivity Commercial magnetometer, or magnetometer used in the heart magnetometry device to detect heart magnetometry, with sensitivity comparable to commercial level, achieving 10fT / Hz -1 / 2 Even lower noise shielding buckets' noise floors can be evaluated.

[0005] The technical solution of the present invention is as follows:

[0006] A method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis, characterized in that it comprises the following steps:

[0007] Step 1: using a first atomic magnetometer to establish a first noise data acquisition channel through the center of the magnetic shielding barrel and form discrete voltage noise data of the first channel, and at the same time using a second atomic magnetometer to establish a second noise data acquisition channel through the center of the magnetic shielding barrel and form discrete voltage noise data of the second channel;

[0008] Step 2, performing Hanning window processing on the discrete voltage noise data of the first channel and the discrete voltage noise data of the second channel respectively;

[0009] Step 3: Perform discrete Fourier transform on the first channel windowed data and the second channel windowed data respectively to obtain the first channel spectrum data V1 (f n ), and the second channel spectrum data V2(f n );

[0010] Step 4: V1(f n ) and V2(f n ) to perform cross-spectral calculation and obtain the power spectrum density X(f n ), using V1(f n ) to obtain the power spectrum density P1(f n ), using V2(f n ) to obtain the power spectrum density P2(f n ), using X(f n ) to obtain the noise floor of the magnetically shielded barrel.

[0011] The first atomic magnetometer and the second atomic magnetometer in step 1 are both 30fT / Hz -1 / 2 Sensitivity Commercial magnetometers, or magnetometers used to detect cardiac magnetism in cardiac magnetometry devices, the photocurrent output by the magnetometer is converted into a voltage through a transimpedance amplifier, and the discrete voltage data is collected through an analog-to-digital conversion collector.

[0012] Step 2 includes the following formula:

[0013]

[0014] Where W H (m) is the window function, M is the total number of sampling points of a single magnetometer, M is a positive integer, m is the sampling point sequence 0≤m≤M, m is an integer, and W is H (m) is multiplied with the discrete voltage data to complete the windowing processing of the magnetometer discrete voltage data.

[0015] Step 3 includes the following formula:

[0016]

[0017] Where v1(m) is the discrete voltage noise data of the first channel, v2(m) is the discrete voltage noise data of the second channel, n is a positive integer, e is a natural constant, and i is an imaginary unit.

[0018] Step 4 includes the following formula:

[0019]

[0020] where f s is the sampling rate, It is the conjugate operation performed on V1(fn).

[0021] Step 4 includes: using the magnetic field voltage response coefficient of the magnetometer for conversion, and calculating the noise floor of the magnetic shielding barrel.

[0022] Step 4 includes the following steps:

[0023] Step 41, retain X(f n ), which is the real part of X(f n )’s amplitude;

[0024] Step 42: For P1(f n ) and P2(f n ) and X(f n ), the average value within each 1Hz frequency range is taken as the power spectrum density value P1, P2 and X at each integer frequency point, P1 corresponds to P1(f n ), P2 corresponds to P2(f n ), X corresponds to X(f n );

[0025] Step 43: Take the square root of X to get the voltage noise amplitude. The unit of voltage noise amplitude is V / Hz -1 / 2 , using the voltage-magnetic field conversion coefficient K of the magnetometer, where the unit of K is V / nT, dividing K by the voltage noise amplitude will give the noise floor value of the magnetic shielding barrel. The larger the value, the higher the noise floor.

[0026] The sensitive axes of the first atomic magnetometer and the second atomic magnetometer in step 1 are parallel, the intrinsic noises of the two magnetometers are non-correlated signals, and the responsiveness of the two magnetometers differs by no more than one order of magnitude.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The highest sensitivity of commercial atomic magnetometers is 30fT / Hz -1 / 2, it is difficult to directly evaluate the noise of the shielded barrel with lower noise, so it is necessary to develop a more accurate metrological magnetometer, and the existing more accurate magnetometers are still in the laboratory stage. This method only requires two independent commercial atomic magnetometers, or any adjacent channels in the cardio-brain magnetic device, to evaluate the noise floor of the shielded barrel through an algorithm. It is low-cost and does not require the complication of the original cardio-brain magnetic device. Furthermore, the operating environment of the cardio-brain magnetic device can also be tested and evaluated to assist in fault diagnosis and analyze the main interference sources of the cardio-brain magnetic device. In addition, this method is essentially to extract the common-mode components between different sensors, and can also be applied to scientific research or commercial applications such as common-mode interference suppression.

[0029] The present invention is based on the Wiener-Hinchin theorem: the power spectral density of any wide-sense stationary random process with a constant mean is the Fourier transform of its autocorrelation function. The dual-channel magnetometer signals are cross-correlated and the results are Fourier transformed. By utilizing the characteristics that the common-mode coherent component between two independent magnetometers is the magnetic noise of the magnetic shielding barrel, while the differential-mode component between channels is incoherent, the autocorrelation of the shielding barrel noise floor signal in the two channels can be obtained by deduction, so the common-mode component of the two independent magnetometers, that is, the noise floor of the magnetic shielding barrel, is obtained through the above cross-spectral operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic flow chart of a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to the present invention. Figure 1 The method comprises the steps of: using a first atomic magnetometer to establish a first noise data acquisition channel through the center of a magnetic shielding barrel and form discrete voltage noise data of the first channel, and using a second atomic magnetometer to establish a second noise data acquisition channel through the center of the magnetic shielding barrel and form discrete voltage noise data of the second channel; step 2, performing Hanning windowing processing on the discrete voltage noise data of the first channel and the discrete voltage noise data of the second channel respectively; and step 3, performing discrete Fourier transform on the windowed data of the first channel and the windowed data of the second channel respectively to obtain the spectrum data V1(f n ), and the second channel spectrum data V2(f n ); Step 4, V1(f n ) and V2(f n ) to perform cross-spectral calculation and obtain the power spectrum density X(f n ), using V1(f n ) to obtain the power spectrum density P1(f n ), using V2(f n ) to obtain the power spectrum density P2(f n ), using X(f n) to obtain the noise floor of the magnetically shielded barrel.

[0031] Figure 2 It is a schematic diagram of the results of evaluating the noise floor of the magnetic shielding barrel by the magnetometer cross-spectrum analysis obtained by implementing the present invention. Figure 2 The horizontal axis is the frequency (Hz, scale value is 0, 10, ···, 90, 100), and the vertical axis is the magnetic noise spectral density (fT / Hz -1 / 2 ), the scale value is 10 0 ,10 1 ,···,10 4 ,10 5 ). Figure 2 The plus sign line in the middle corresponds to the measurement of the first atomic magnetometer, the dot line in the middle corresponds to the measurement of the second atomic magnetometer, and the circle line in the middle corresponds to the noise floor of the magnetic shielding barrel extracted by the cross-spectrum. Figure 2 Medium 8.6fT / Hz -1 / 2 It is the noise floor of the magnetic shielded barrel at 30Hz frequency.

[0032] Figure 3 This is a schematic diagram of the results of measuring the noise floor of a magnetic shielding barrel using a high-sensitivity dual-light magnetometer. Figure 3 Medium 8.6fT / Hz -1 / 2 is the magnetic noise at 30Hz, and Figure 2 This proves the effectiveness of the method of the present invention. DETAILED DESCRIPTION

[0033] Below is the attached figure ( Figure 1-Figure 3 ) and Examples illustrate the present invention.

[0034] Figure 1 It is a schematic flow chart of a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to the present invention. Figure 2 It is a schematic diagram of the results of evaluating the noise floor of the magnetic shielding barrel by the magnetometer cross-spectrum analysis obtained by implementing the present invention. Figure 3 This is a schematic diagram of the results of measuring the noise floor of a magnetic shielding barrel using a high-sensitivity dual-light magnetometer. Figures 1 to 3 As shown, a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis includes the following steps: step 1, using a first atomic magnetometer to establish a first noise data acquisition channel through the center position of the magnetic shielding barrel and form discrete voltage noise data of the first channel, and at the same time using a second atomic magnetometer to establish a second noise data acquisition channel through the center position of the magnetic shielding barrel and form discrete voltage noise data of the second channel; step 2, performing Hanning windowing processing on the discrete voltage noise data of the first channel and the discrete voltage noise data of the second channel respectively; step 3, performing discrete Fourier transform on the windowed data of the first channel and the windowed data of the second channel respectively, and obtaining the spectrum data V1(f n ), and the second channel spectrum data V2(fn ); Step 4, V1(f n ) and V2(f n ) to perform cross-spectral calculation and obtain the power spectrum density X(f n ), using V1(f n ) to obtain the power spectral density p1(f n ), using V2(f n ) to obtain the power spectrum density P2(f n ), using X(f n ) to obtain the noise floor of the magnetically shielded barrel.

[0035] The first atomic magnetometer and the second atomic magnetometer in step 1 are both 30fT / Hz -1 / 2 Sensitivity Commercial magnetometers, or magnetometers used to detect cardiac magnetism in cardiac magnetometry devices, the photocurrent output by the magnetometer is converted into a voltage through a transimpedance amplifier, and the discrete voltage data is collected through an analog-to-digital conversion collector.

[0036] Step 2 includes the following formula:

[0037]

[0038] Where W H (m) is the window function, M is the total number of sampling points of a single magnetometer, M is a positive integer, m is the sampling point sequence 0≤m≤M, m is an integer, and W is H (m) is multiplied with the discrete voltage data to complete the windowing processing of the magnetometer discrete voltage data.

[0039] Step 3 includes the following formula:

[0040]

[0041]

[0042] Where v1(m) is the discrete voltage noise data of the first channel, v2(m) is the discrete voltage noise data of the second channel, n is a positive integer, e is a natural constant, and i is an imaginary unit.

[0043] Step 4 includes the following formula:

[0044]

[0045] where f s is the sampling rate, It is the conjugate operation performed on V1(fn).

[0046] Step 4 includes: using the magnetic field voltage response coefficient of the magnetometer for conversion, and calculating the noise floor of the magnetic shielding barrel.

[0047] Step 4 includes the following steps: Step 41, retaining X(f n ), which is the real part of X(f n ); Step 42: for p1(f n ) and P2(f n ) and X(f n ), the average value within each 1 Hz frequency range is taken as the power spectrum density value P1, P2 and X at each integer frequency point, P1 corresponds to p1(f n ), P2 corresponds to P2(f n ), X corresponds to X(f n ); Step 43: Take the square root of X to obtain the voltage noise amplitude, the unit of the voltage noise amplitude is V / Hz -1 / 2 , using the voltage-magnetic field conversion coefficient K of the magnetometer, where the unit of K is V / nT, dividing K by the voltage noise amplitude will give the noise floor value of the magnetic shielding barrel. The larger the value, the higher the noise floor.

[0048] The sensitive axes of the first atomic magnetometer and the second atomic magnetometer in step 1 are parallel, the intrinsic noises of the two magnetometers are non-correlated signals, and the responsiveness of the two magnetometers differs by no more than one order of magnitude.

[0049] The present invention discloses a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis, which relates to the field of magnetic field measurement technology and weak signal detection. The method comprises the following steps: data acquisition, using two independent atomic magnetometers to simultaneously collect noise data at the center of the magnetic shielding barrel, using a Hanning window to perform windowing processing on the discretized noise data of the two channels to reduce near-sidelobe leakage, and performing Fourier transform on the windowed dual-channel data to obtain spectrum data, and finally performing frequency domain cross-correlation conjugate operation to calculate the frequency domain characteristics of the common mode components of the two atomic magnetometers, and analyzing to obtain the noise spectrum of the common mode components of the two magnetometers, that is, the noise floor of the magnetic shielding barrel. The present invention uses the data collected by the two-channel magnetometers to perform cross-spectrum analysis operation to extract the common mode components between the channels, thereby evaluating the noise floor of the magnetic shielding barrel, or the noise floor of the magnetic shielding environment, without the need to develop and use a higher-precision metrological magnetometer, and only relying on two magnetometers with lower sensitivity to evaluate the noise floor of the magnetic shielding barrel, which is lower in cost, does not need to increase the equipment complexity of the extremely weak magnetic measurement device, and has high credibility.

[0050] A method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis collects data from atomic magnetometers of two independent channels, and performs windowing function, Fourier transform, and cross-spectrum operation respectively. Finally, the noise floor of the magnetic shielding barrel is measured through the cross-spectrum density map. The method specifically includes the following steps:

[0051] S1: Two independent atomic magnetometers are used to simultaneously collect noise data at the center of the magnetic shielding barrel. The photocurrent output by the magnetometer is converted into a voltage through a transimpedance amplifier, and the discrete voltage data is collected through an analog-to-digital conversion collector;

[0052] S2: The discrete voltage noise data collected from the two channels are windowed using a Hanning window to reduce near-sidelobe leakage;

[0053] S3: Perform discrete Fourier transform on the windowed dual-channel data to obtain spectrum data V1 (f n ) and V2(f n );

[0054] S4: Convert the spectrum data of magnetometers 1 and 2 into power spectrum and divide it by the frequency resolution to obtain the power spectrum density P1 (f n ) and P2(f n ), the power spectrum density is used to analyze the noise of magnetometers 1 and 2. If the power spectrum density obtained in this step is greater than the power spectrum density obtained by the cross spectrum, it means that the noise floor of the shielding barrel is low, and the noise floor of the magnetic shielding barrel is obtained by the cross spectrum. If the power spectrum density obtained in this step is similar to the power spectrum density obtained by the cross spectrum, it means that the current magnetic shielding barrel has a large noise, and the noise floor of the magnetic shielding barrel can be directly obtained through the power spectrum density data of the magnetometer;

[0055] The frequency spectrum V1(f n ) and V2(f n ) to perform cross-spectrum calculation and calculate the power spectrum density data X(f n ), the magnetic field voltage response coefficient of the magnetometer is used for conversion, and the magnetic noise floor of the shielding barrel is calculated.

[0056] The two independent atomic magnetometers refer to atomic magnetometers using different lasers and different analog-to-digital conversion collector collection channels, so that the intrinsic noises of the two magnetometers are non-correlated signals, and the response of the two magnetometers needs to differ by no more than one order of magnitude.

[0057] The analog-to-digital conversion collector collects discrete voltage data, and the sampling rate is downsampled from the analog-to-digital conversion collector sampling rate to 4kSPS (4 thousand sampling points per second). In order to make the power spectrum density obtained later smooth enough, the collection time is required to be more than 30 minutes.

[0058] The Hanning window function expression in step S2 is:

[0059]

[0060] Where M is the total number of sampling points of a single magnetometer, and m is the sampling point sequence 0≤m≤M. H(m) is a window function, which is multiplied by the discrete voltage data to complete the windowing processing of the discrete voltage data of the magnetometer.

[0061] The discrete Fourier expression in step S3 is:

[0062]

[0063] Where v1(m) is the discrete voltage data of magnetometer 1, v2(m) is the discrete voltage data of magnetometer 2, V1(f n ) is the fast Fourier transform spectrum of magnetometer 1, V2(f n ) is the fast Fourier transform spectrum of magnetometer 2, f n is a frequency sequence.

[0064] The power spectrum density expression in step S4 is:

[0065]

[0066] where f s is the sampling rate, P1(f n ) is the power spectral density of the first magnetometer, P2(f n ) is the power spectral density of the second magnetometer.

[0067] The magnetic noise in the magnetic shielding barrel is a wide-sense stationary random process with a constant mean value, and the cross-spectrum expression is:

[0068]

[0069] Among them, X(f n ) is the power spectrum density of the common mode components of the two channel magnetometers obtained by cross-spectral analysis, It is the conjugate operation of V1(f).

[0070] Step S4 further comprises:

[0071] S41: Keep X(f n ), which is the magnitude of the power spectral density;

[0072] S42: The power spectrum density P1 (f n ) and the power spectral density P2(f n ), and X(f n ), and take the average value within each 1 Hz frequency range as the power spectrum density value of the corresponding integer frequency point;

[0073] S43: Calculate the voltage noise amplitude (unit: V / Hz) by changing the result of S42. -1 / 2), the voltage-to-magnetic-field conversion coefficient of the magnetometer, also called the responsivity K (unit: V / nT), is divided by the voltage noise amplitude to calculate the magnetic noise floor in the shielding barrel. The larger the value, the higher the noise floor.

[0074] To facilitate the implementation of the present invention by ordinary technicians in the field, the present invention takes a magnetically enclosed shielding barrel with an outer diameter of 45 cm, a length of 90 cm, 4 layers of permalloy inside and an outermost layer of aluminum as an example to illustrate the implementation process of the present invention. The shielding barrel has been demagnetized and is a magnetic shielding barrel configuration commonly used in laboratories. The present invention provides a method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis, comprising the following steps:

[0075] S1: Using two independent atomic magnetometers, the atomic source uses 87 Rb, heated to 120 degrees Celsius, and compensated for the static magnetic field around the magnetometer. The magnetometer uses a 795nm laser to complete atomic pumping, which is used to polarize rubidium alkali metal atoms. At the same time, a photoelectric converter is used to convert the light intensity of the pumping light after passing through the atomic gas chamber into a change in photocurrent, and then converted into a voltage signal for collection through a transimpedance amplifier in the electrical acquisition module. A 1kHz modulated magnetic field is applied in a direction perpendicular to the plane of the pumping light direction, and the direction in which the modulated magnetic field is applied is the sensitive direction of the atomic magnetometer. Here, the sensitive direction is set to the axial direction of the magnetic shielding barrel. The present invention requires that the sensitive axis directions of the two magnetometers are parallel, and independence refers to the two magnetometers using different lasers and different analog-to-digital conversion collectors to collect channels of atomic magnetometers, so that the intrinsic noise of the two magnetometers is a non-correlated signal, and the two magnetometers need to differ in responsivity by no more than one order of magnitude. Further, the analog-to-digital conversion collector collects discrete voltage data, and the sampling rate is downsampled to 4kSPS from the sampling rate of the analog-to-digital conversion collector. In order to make the power spectrum density obtained later smooth enough, the acquisition time is required to be more than 30 minutes. The collected voltage signal is demodulated at 1kHz through the phase-locked amplifier module in the electrical acquisition module to obtain the discrete voltage data v1(m) and v2(m) of the two atomic magnetometers. In addition, the two magnetometers need to perform synchronous data acquisition, which is very important for the subsequent cross-spectrum analysis to extract the common mode component.

[0076] S2: The discrete voltage noise data collected from the two channels are windowed using a Hanning window to reduce near sidelobe leakage.

[0077] In order to better reduce the interference signal adjacent to the main lobe in the spectrum, that is, the near side lobe leakage, the Hanning window is used. The window function coefficients applied use the root mean square value, not the maximum value, so choose The window function expression is:

[0078]

[0079] Where M is the total number of sampling points of a single magnetometer, and m is the sampling point sequence 0≤m≤M. H (m) is a window function, which is multiplied by the discrete voltage data to complete the windowing processing of the discrete voltage data of the magnetometer.

[0080] S3: Perform discrete Fourier transform on the windowed dual-channel data to obtain spectrum data V1 (f n ) and V2(f n );

[0081] The discrete Fourier expression in this step is:

[0082]

[0083] S4: Convert the spectrum data of magnetometers 1 and 2 into power spectrum and divide by the frequency resolution M*f s , the power spectral density diagram of the two-channel magnetometer is obtained, and the noise spectrum of each channel is evaluated to compare the results of S5 cross-spectral analysis to verify the correctness of common mode extraction;

[0084] The power spectral density expression in this step is:

[0085]

[0086] where f s is the sampling rate, P1(f n ) is the power spectral density of the first magnetometer, P2(f n ) is the power spectral density of the second magnetometer.

[0087] The frequency spectra of the two magnetometers are cross-spectral operated to calculate the frequency domain characteristics of the common mode components of the two atomic magnetometers. The common mode component is the magnetic noise floor, that is, the magnetic noise of the magnetic shielding barrel. The shielding barrel noise floor is calculated through the response coefficient of the magnetometer.

[0088] The cross-spectral expression is:

[0089]

[0090] Among them, X(f n) is the power spectrum of the common mode components of the two channel magnetometers obtained by cross-spectral analysis. The condition for the validity of this formula is to assume that the response voltage v1(t) collected by the analog-to-digital converter of magnetometer 1 and the response voltage v2(t) collected by the analog-to-digital converter of magnetometer 2 have the characteristics: v1(t)=s(t)+n1(t) and v2(t)=s(t)+n2(t), where s(t) is the common mode component of the two magnetometers, that is, the magnetic noise of the operating environment of the magnetometer, n1(t) is the uncorrelated background noise of magnetometer 1 and the uncorrelated electrical noise introduced in the signal processing process, and n2(t) is the uncorrelated background noise of magnetometer 2 and the uncorrelated electrical noise introduced in the signal processing process. Therefore, when the dual channels are cross-correlated, only the term s(t)+s(t-τ), that is, the autocorrelation term of the shielding barrel magnetic noise is retained, and the other terms are 0, where s(t-τ) is the delay τ of s(t). Therefore, the result is subjected to Fourier transform, and the order of integration is exchanged. By using the time-shift property of Fourier transform, the expression of the above shielding barrel noise power spectrum can be obtained. Furthermore, it is also necessary to convert the power spectrum into units for evaluating the noise floor of the magnetic shielding barrel. Therefore, step S4 further includes:

[0091] S41: Keep X(f n ), which is the magnitude of the power spectral density;

[0092] S42: The power spectrum density P1 (f n ) and the power spectral density P2(f n ), and X(f n ), take the average value within each 1 Hz frequency range as the power spectrum density value at the integer frequency point;

[0093] S43: Calculate the voltage noise amplitude (unit: V / Hz) by changing the result of S42. -1 / 2 ), the voltage-to-magnetic-field conversion coefficient of the magnetometer, also called the responsivity K (unit: V / nT), is divided by the voltage noise amplitude to calculate the magnetic noise floor in the shielding barrel. The larger the value, the higher the noise floor.

[0094] Finally, the cross-spectrum analysis results of the noise floor of the shielded magnetic shielding barrel are as follows Figure 2 As shown in the figure, magnetometers 1 and 2 synchronously collect 60 minutes of 4kSPS magnetic shielding barrel noise data, which does not contain any signal. Since the magnetic shielding barrel for biomagnetism is based on the characteristics of biological signals, it usually focuses on the frequency range of 2 to 100Hz. Therefore, this example evaluates the noise floor of the magnetic shielding barrel at a frequency point near 30Hz. As shown by the dotted line in the figure, the magnetometer evaluated by the method of the present invention is 8.6fT / Hz -1 / 2Furthermore, in order to verify the correctness of the evaluation results, a high-sensitivity dual-light magnetometer with a 10mm air chamber developed in the laboratory was used, and a calibrated coil was used in a magnetic shielding environment to apply a 30Hz, 100pT average calibration magnetic field to obtain the magnetometer response coefficient K. Figure 3 There are peaks at 30Hz and 60Hz. Figure 3 The magnetic noise measured at the same position is also 8.6fT / Hz -1 / 2 , which proves the effectiveness of the method of the present invention.

[0095] The above process belongs to offline data processing. After data collection, it is necessary to use a computer to process the data. Each step is completed through the response function of Matlab to evaluate the noise floor of the magnetic shielding barrel. It can also be described in C language and burned into the array magnetometer control system through the compilation software. It is used to evaluate the operating environment noise of the array atomic magnetometer device and perform fault analysis.

[0096] The present invention is based on the Wiener-Hinchin theorem: the power spectral density of any generalized stationary random process with a constant mean is the Fourier transform of its autocorrelation function. By processing cross-spectral data, two magnetometers are used to complete the noise floor evaluation of the magnetic shielding barrel with lower noise than the magnetometer background, which is lower in cost and simple to operate. This method is also applicable to different magnetic shielding spaces and can be applied to other fields to extract common mode components for analysis, and has high practical value in engineering applications.

[0097] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in the field. It is pointed out here that the above description helps those skilled in the art to understand the invention, but does not limit the protection scope of the invention. Any equivalent replacement, modification and / or simplification of the above description without departing from the essence of the invention falls within the protection scope of the invention.

Claims

1. A method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis, characterized in that: The following steps are involved: Step 1: using a first atomic magnetometer to establish a first noise data acquisition channel through the center of the magnetic shielding barrel and form discrete voltage noise data of the first channel, and at the same time using a second atomic magnetometer to establish a second noise data acquisition channel through the center of the magnetic shielding barrel and form discrete voltage noise data of the second channel; Step 2, performing Hanning window processing on the discrete voltage noise data of the first channel and the discrete voltage noise data of the second channel respectively; Step 3: Perform discrete Fourier transform on the first channel windowed data and the second channel windowed data respectively to obtain the first channel spectrum data V1 (f n ), and the second channel spectrum data V2(f n ); Step 4: V1(f n ) and V2(f n ) to perform cross-spectral calculation and obtain the power spectrum density X(f n ), using V1(f n ) to obtain the power spectrum density P1(f n ), using V2(f n ) to obtain the power spectrum density P2(f n ), using X(f n ) to obtain the noise floor of the magnetically shielded barrel.

2. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: The first atomic magnetometer and the second atomic magnetometer in step 1 are both 30fT / Hz -1 / 2 Sensitivity Commercial magnetometers, or magnetometers used to detect cardiac magnetism in cardiac magnetometry devices, the photocurrent output by the magnetometer is converted into a voltage through a transimpedance amplifier, and the discrete voltage data is collected through an analog-to-digital conversion collector.

3. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: Step 2 includes the following formula: Where W H (m) is the window function, M is the total number of sampling points of a single magnetometer, M is a positive integer, m is the sampling point sequence 0≤m≤M, m is an integer, and W is H (m) is multiplied with the discrete voltage data to complete the windowing processing of the magnetometer discrete voltage data.

4. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: Step 3 includes the following formula: Where v1(m) is the discrete voltage noise data of the first channel, v2(m) is the discrete voltage noise data of the second channel, n is a positive integer, e is a natural constant, and i is an imaginary unit.

5. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: Step 4 includes the following formula: where f s is the sampling rate, It is the conjugate operation performed on V1(fn).

6. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: Step 4 includes: using the magnetic field voltage response coefficient of the magnetometer for conversion, and calculating the noise floor of the magnetic shielding barrel. Step 4 includes the following steps: Step 41, retain X(f n ), which is the real part of X(f n )’s amplitude; Step 42: For P1(f n ) and P2(f n ) and X(f n ), the average value within each 1Hz frequency range is taken as the power spectrum density value P1, P2 and X at each integer frequency point, P1 corresponds to P1(f n ), P2 corresponds to P2(f n ), X corresponds to X(f n ); Step 43: Take the square root of X to get the voltage noise amplitude. The unit of voltage noise amplitude is V / Hz -1 / 2 , using the voltage-magnetic field conversion coefficient K of the magnetometer, where the unit of K is V / nT, dividing K by the voltage noise amplitude will give the noise floor value of the magnetic shielding barrel. The larger the value, the higher the noise floor.

7. The method for evaluating the noise floor of a magnetic shielding barrel based on cross-spectrum analysis according to claim 1, characterized in that: The sensitive axes of the first atomic magnetometer and the second atomic magnetometer in step 1 are parallel, the intrinsic noises of the two magnetometers are non-correlated signals, and the responsiveness of the two magnetometers differs by no more than one order of magnitude.

Images